Large language models revive detailed legacy drug study reports quickly
From archive to resource: bringing legacy data back to life at scale with large language models
Information Retrieval
Summary
Old pharmaceutical study reports contain important details often lost in current data systems, making it hard to reuse the data. The authors developed smart computer programs that use large language models to automatically pull out and organize this hidden information from many old reports. These programs work much faster than humans and provide clear evidence to help verify results. They showed that recovered details, like whether animals were fasting, can explain differences in medical test results—helping scientists use old data more effectively in new studies.
What this means in practice
- •For pharmaceutical data teams: Automate extraction of detailed experimental conditions from legacy study reports to improve data reuse and comparability.
- •For healthcare data integrators: Integrate recovered preclinical metadata into clinical data systems to enhance historical control selection and analysis accuracy.
- •For regulatory affairs professionals: Use structured and traceable legacy study metadata to support regulatory submissions by providing clearer experimental context.$Commercial implications: Creates software tools that improve regulatory dossier quality and speed by structuring hidden experimental metadata.
Authors
C. H. Vieira-Vieira, I. Schneider Chernov, A. Serra, M. Correa Marrero, P. L. Bhavani Mohan, G. Duchateau-Nguyen, T. Y. Doktorova, A. Kreuchwig
Abstract
Pharmaceutical companies maintain vast archives of legacy preclinical study reports that contain study-context details, such as fasting, anesthesia, and laboratory instrumentation, that are rarely captured in structured data formats, yet they can correlate with clinical pathology parameters values and limit the reuse of historical data. This work demonstrates practical, scalable solutions to recover such descriptors from heterogeneous report texts using large language model (LLM)-based extraction pipelines developed independently by two industry teams. The pipelines identify and normalize metadata fields and return verbatim evidence with page and section reference, hasting human verification and iterative refinement. Evaluation of over 400 documents against expert-curated reference values showed high agreement across a broad set of metadata fields. Performance was strongest for consistently reported items and lower for descriptors subject to linguistic or report structure variability. Extraction time has shrunk from 25 minutes in the manual benchmark to 1.6 minutes. As a concrete biological demonstration, we show that fasting status extracted from report text was associated with species- and analyte-specific shifts in clinical chemistry parameters, illustrating how recovered metadata can improve selection of historical controls. Key strengths of the approach are scalability, traceability, and the ability to recover information missing from current industry standards. Main limitations are dependence on source-report quality and initial expert effort to define and refine schemas. These results establish how LLMs can convert heterogeneous, unstructured legacy reports into evidence-grounded, reusable and structured metadata, leading to better experimental comparability, and thus enabling reliable data pooling or informed virtual control selection.